University of Sydney researchers have released two studies examining the resources used to produce academic research and the financial returns of major publishing groups, reopening questions about who pays for scholarly communication.
The university announced the findings on October 15. Both papers are preprints, meaning they have been made public before peer-reviewed journal publication. Their estimates should be assessed with that limitation in mind.
The work was supported through a University of Sydney Industry Partnership Grant with SPARC, an organization that advocates open access to research. That partnership is relevant context for the studies’ focus.
What it costs to produce a paper
One study, The Cost of Knowledge, models the resources used to produce a research article before publication. These include academic labor and institutional support, rather than simply a publisher’s production expenses.
Its illustrative case involves a mid-career academic at a US public doctoral university producing a social-science paper. The modeled cost is about $70,000.
That figure is not an observed average for all published research. A laboratory experiment, a clinical trial and a social-science article can require very different resources. Costs also depend on salaries, institutional arrangements and how researchers’ time is assigned.
The model nevertheless raises a useful accounting question: a journal’s publication charge is only one component of the cost of making research available. Universities and funders have already supported much of the work that creates the article.
Publisher returns need the right boundary
The companion preprint, How Profitable Are the Big Four Publishers?, examines RELX, Informa, Springer Nature and Wiley over 2005 to 2025.
The researchers calculate $38.8 billion returned to shareholders over that period. These are returns from the parent groups, whose activities extend beyond academic journal articles. They should not be described as money derived entirely from publicly funded research.
Dividends distribute cash to shareholders. Share repurchases use company funds to buy back shares. Neither is the same accounting measure as revenue or operating profit, and totals across two decades do not show what a publisher earned on an individual paper.
Research publishing also includes tools and services
RELX’s official 2025 results describe its Scientific, Technical & Medical business as including databases, tools, electronic reference products and primary research.
The company attributes growth partly to analytics and decision tools, including AI-enabled products. That product mix matters when assessing claims about the profitability of publishing: a business segment can earn money from services beyond the sale of access to journal articles.
For universities negotiating research-access agreements, the relevant comparison is therefore more detailed than a single profit figure. It includes the content and services purchased, the rights retained by authors and the separate cost of making papers openly available.
The two preprints provide material for that debate, but neither establishes one universal cost for research or a price that every institution should pay. Their assumptions, institutional examples and financial boundaries need to remain visible when their conclusions are used in funding or procurement decisions.
